Download CORALIE-ELODIE new planets and planetary systems. Looking for

Planetary Systems in the Universe - Observation, Formation and Evolution
Proceedings [AU Symposium No. 202, @2004 [AU
Alan Penny, Pawel Ariymowicz, Anne-Marie Lagrange, & Sara Russell, eds.
CORALIE-ELODIE new planets and planetary systems.
Looking for fossil traces of formation and evolution
S. Udry, M. Mayor and D. Queloz
Geneva Observatory, CH-1290 Sauvemy, Switzerland
Abstract.
6 new extra-solar planet candidates ( HD 6434 b, HD 19994 b,
HD 83443c, HD 92788 b, HD 121504 b, HD 190228 b ) are announced as part
of our planet-search programmes in the northern and southern hemispheres.
HD 83443 c is member of a 2-planet system with Saturnian and sub-Saturnian
masses. Another system including a planet + a very low-mass brown dwarf orbiting HD 168443 is also presented. These 2 new systems and the new planetary
detections rise to 25 the number of ELODIE and CORALIE candidates with
minimum masses ~ 20MJup. The orbital element distributions of giant-planet
candidates, like the secondary mass function, the eccentricity and period distributions, compared to the equivalent distributions for spectroscopic binaries,
strongly suggest different formation mechanisms for the two populations.
1.
Introduction
New enthusiastic results in the domain of extra-solar planets are popping up on
an almost daily basis. From one meeting to the next, not only new candidates
are announced by the teams of "planet hunters" monitoring high-precision radial
velocities (RV) but also number of various approaches to the field are explored,
trying to better understand planetary formation as new observational results
often tend to set new questions rather than bring definitive answers to the variety
of explanations proposed for the observed properties of these systems.
In this contribution new discoveries are presented and the global properties
of the extra-solar planet sample as a whole are discussed as well. The first part of
the paper will be dedicated to the description of new planet candidates obtained
in the frame of two large planet-search programmes around solar-type stars,
in which our group in Geneva is involved, namely the ELODIE high-precision
RV survey in Haute-Provence (France) for the northern sky and the CORALIE
planet-search programme at ESO-LaSilla (Chile) in the south.
Among the new ELODIE results we will focus on the spectroscopic transit of the planet in front of HD 209458 and on the new planetary companion
orbiting the star HD 190228. The CORALIE programme is rich in new prominent results with 5 new planets around HD 6434, HD 19994, HD 83443, HD 92788,
and HD 121504. HD 83443 turns out to be the 2nd known exoplanetary system,
hosting two very low-mass planets, respectively close to and of half the mass of
Saturn. An additional 3rd new system composed of a massive planet + a very
low-mass brown dwarf (15.1 MJup/ sin i) orbiting HD 168443 is presented as well.
Finally, the end part of this contribution will be devoted to a global description of the extra-solar planets properties in terms of orbital element distributions
12
Downloaded from https:/www.cambridge.org/core. IP address: 88.99.165.207, on 18 Jun 2017 at 00:26:26, subject to the Cambridge Core terms of
use, available at https:/www.cambridge.org/core/terms. https://doi.org/10.1017/S0074180900217403
13
CORALIE-ELODIE new planets and planetary systems
Table 1.
Sp
V
MI
P
T
e
V
w
K
m2
a
N
RMS
a
Stellar and orbital parameters for
14Herb
KOV
6.6
0.79
1787.5 ± 16.5
49579.3 ± 23.3
0.34 ± 0.02
-13.842
22.05 ± 3.89
91 ± 2
4.2
2.68
HD 168443 b
58.117 ±
51558.36 ±
0.526 ±
172.2 ±
473 ±
7.2
0.29
b
CORALIE,
and HD 168443
HD 168443c
G5
6.9
0.92
0.014
1667 ±
0.12
50269.5 ±
0.265 ±
0.008
-48.744
1.1
59 ±
6
288 ±
15.1
2.67
48
36
0.049
5
13
58 b + 30 c
7.6
84 a
12.7
ELODIE,
14 Her
c
Units
[mag]
[M0]
[days]
[JD-2400000]
[kms- I ]
[deg]
[ms- I ]
[MJup/ sin i]
[AU]
[ms- I ]
HIRES (from Marcy et al. 1999b)
emphasizing their differences from the equivalent distributions for stellar companions to solar-type stars.
2.
The ELODIE high-precision radial-velocity survey
Our RV survey in the northern hemisphere to search for planets orbiting solartype stars is a Swiss-French collaboration including the Geneva, the Grenoble
and the Haute-Provence observatories. The high-precision RVs are measured
by numerically cross-correlating the ELODIE stellar spectra with a solar-type
template (Baranne et al. 1996). ELODIE is a two-fiber fed high-resolution
echelle spectrograph (R ~ 42000) mounted on the 193-cm telescope of the
Haute-Provence Observatory. The two fibers are used to simultaneously take
the stellar spectra and follow with a ThAr calibration lamp the instrumental
drifts due to environmental effects (temperature and pressure changes). After
the introduction in the fiber paths of double-scrambling devices, the achieved
precision for RV measurements is now of 8-10 ms- I .
The ELODIE planet-search sample is magnitude-limited and consists of a
350 F-K dwarfs of the solar vicinity. Its is an extension of the original MayorQueloz sample out of which 51 Peg b, the first extra-solar planet around a solartype star, was discovered (Mayor & Queloz 1995). In addition to this very
mediatic early success of the programme, 4 additional planetary companions
have been detected with ELODIE:
- G1876 b (Delfosse et al. 1999; also found by Marcy et al. 1999a) orbiting
an M4 dwarf and discovered as part of a binarity study of M dwarfs.
- 14 Her b; the long-period orbit is completely covered now. Updated orbital elements and RV curves are given in Table 1 and Fig. 1 (left).
- HD 209458 b; ELODIE and CORALIE measurements combined with HIRES
data obtained at Keck 1 by D. Latham et al. allowed us to precisely determine
the epoch of the planetary transit (Mazeh et al. 2000) then used by Charbonneau et al. (2000) to observed the photometric transit twice in September 1999.
The transit was also independently observed by Henry et al. (2000). The phoI'V
Downloaded from https:/www.cambridge.org/core. IP address: 88.99.165.207, on 18 Jun 2017 at 00:26:26, subject to the Cambridge Core terms of
use, available at https:/www.cambridge.org/core/terms. https://doi.org/10.1017/S0074180900217403
14
Udry, Mayor and Queloz
Figure 1.
Left. RVs and updated Keplerian solution for 14 Her. Right,
ELODIE spectroscopic transit of HD 209458: RVs and best solution for the
anomaly curve corrected from the orbital motion (Queloz et al. 2000). The
1 a confidence level is indicated by the hatched area
tometric transits brought the proof of the gaseous nature of HD 209458 (Mazeh
et aI. 2000). Further results on the geometry of the system are obtained by the
spectroscopic observation of the transit with ELODIE (next section).
- The new massive planet (m2 sin i = 5 MJup) orbiting HD 190228 on a
long-period (P = 1161 d), eccentric (e = 0.5) orbit. The somehow large distance of the star from the Sun (d = 62 pc) prevents the Hipparcos astrometric
data to bring definitive constraints on the real mass of the planetary companion. Further details on the primary star, orbital elements and inferred planetary
characteristics as well as the RV curve are given in Sivan et aI. (this volume).
2.1.
HD 209458: spectroscopic transit
The star HD 209458 became famous end of 1999 when the transit of a planet in
front of the stellar disk was observed (Charbonneau et aI. 2000, Henry et aI.
2000), providing us for the first time with physical characteristics of an exoplanet
(Mazeh et aI. 2000). Observations with better facilities are now drastically
improving these early results (see e.g. the impressive transit luminosity curve
obtained with the HST by Brown et aI., this volume). Here, we will concentrate
on the transit observed spectroscopically with ELODIE in November 1999.
In the same way as spots, the shadow on the stellar disk of a transiting
planet induces deformations of the spectral lines by hidding part of the light
coming from the approaching or receding sides of the star. To see these deformations on individual lines requires very high-resolution and high signal-to-noise
spectra whereas they can be easily observed on our cross-correlation functions
which then are slightly red- or blue-shifted. This effect influences the measured
RVs in the form of an anomaly of the orbital curve (Fig. 1, right) whose shape
depends on the geometry of the transit and on stellar parameters. If the orbital
and the apparent stellar equatorial planes are parallel, the lobes of the anomaly
curve are symmetric, also the larger the stellar rotation, the larger the amplitude
Downloaded from https:/www.cambridge.org/core. IP address: 88.99.165.207, on 18 Jun 2017 at 00:26:26, subject to the Cambridge Core terms of
use, available at https:/www.cambridge.org/core/terms. https://doi.org/10.1017/S0074180900217403
CORALIE-ELODIE new planets and planetary systems
15
of the observed anomaly (Eggenberger 2000, master thesis; Queloz et al. 2000).
The main results can be summarized as follow (Fig. 1, right): 1) the positive sign
of the anomaly curve during the ingress indicates that the stellar rotation is in
the same direction as the planet motion; 2) an angle ti. L. ~ 3.9 0 , but with large
uncertainty, is estimated between the rotation and orbital axes; 3) the method
provides an independent estimate of the stellar rotation: v sin i == 3.75 kms- I.
3.
The CORALIE planet-search programme: new results
Since the summer of 1998, a large high-precision RV survey is being carried out
by our Geneva group with the CORALIE echelle spectrograph on the 1.2-m Euler
Swiss telescope at ESO- La Silla. CORALIE is the little sister of ELODIE with
improved resolution (R ~ 50000) and smaller CCD pixels which allow to reach,
with the same cross-correlation technique, a precision of 5-7 ms- I.
The CORALIE planet-search sample consists of more than 1650 F8 to MO
solar-type stars selected according to distance in order to obtain a statistically
well-defined volume-limited set of dwarfs of the solar neighbourhood (Udry et
al. 2000). In addition to the planet search, this sample will allow to collect
information on spectroscopic binaries and give us a synthetic view on companions
to solar-type stars from q == m2/mI == 1 down to q :S 0.001.
The CORALIE programme is very fruitful. In about 2 years, ·it has already revealed 19 low-mass companions with m2 sin i :S 16 MJup, among which
16 have minimum masses well in the planet domain (m2 sin i < 5 MJuP). As
measurements are pilling up, the detection frequency is rising: 15 among the
19 candidates were announced between January and August 2000 (Naef et al.
2000; ESO 2000; this conference). From the observed frequency of giant planets,
we expect about 100 planets out of the CORALIE sample.
In addition to the new ELODIE candidate, five new planets and one very
low-mass brown dwarf detected with CORALIE are announced. Two (HD 83443 c
and HD 168443 c) of them are part of multi-component systems and they are
described below, and four are additional planetary candidates orbit the stars
HD6434, HD 19994, HD92788 (also presented by Butler et al. 2000, this volume) and HD 121504. The corresponding main stellar and orbital characteristics as well as inferred planetary properties and RV curves are given in Queloz
et al. (this volume). In summary, 2 candidates have "intermediate" periods
and low minimum masses (HD 6434: m2 sin i == 0.48 MJup, P == 22.1 d; HD 121504:
m2sini==0.89MJup, P==64.6d), and two heavier planets are on long-period orbits (HD 19994:
P == 454 d, m2 sin i == 2.0 MJup:
HD 92788, P == 340 d,
m2 sin i == 3.8 M Jup). The new candidate characteristics are very similar to properties of previously detected planets. In particular, the orbits are eccentric when
the period is not too short.
3.1.
HD 83443: a 2-planet system with Saturn and sub-Saturn masses
The most interesting new CORALIE candidate is the 2-planet system orbiting HD 83443. The short-period planet was already announced a few months
ago in another set of 8 CORALIE new discoveries (ESO 2000). The period
P == 2.985 days is very close to 3 sidereal days. The planet has the shortest peDownloaded from https:/www.cambridge.org/core. IP address: 88.99.165.207, on 18 Jun 2017 at 00:26:26, subject to the Cambridge Core terms of
use, available at https:/www.cambridge.org/core/terms. https://doi.org/10.1017/S0074180900217403
16
Udry, Mayor and Queloz
Figure 2.
Simultaneously-derived orbital solutions for the HD 168443 system. The short (top left, b) and long (bottom left, c) periods are shown
individually. The combined orbit is displayed as well (top right). Lower right.
Stability analysis: temporal evolution over 105 years of the eccentricity and
separation parameters for both (b and c) orbits
riod and the smallest separation to the parent star among the known extra-solar
planets. Early on, an abnormal behavour was detected in the residuals of the
short-period Keplerian solution showing a periodicity of P == 29.83 days (almost
exactly 10 times the inner period). Very low masses are inferred for the planets,
respectively close to and of half the mass of Saturn.
Looking at such a system with 2 planets on so close orbits immediately rises
concerns about the system dynamical stability. A preliminary analysis clearly
confirms the system stability. Nevertheless dynamical perturbations between
the planets are far from being negligible and in particular may explain the noncircularity of the inner orbit despite of its very short period. Further orbital
details and comments are given in Mayor et al. (this volume).
3.2.
HD 168443, a 3rd system: a planet
+a
low-mass brown dwarf
Another intriguing system, composed of a heavy giant planet of rv 7.2 M Jup/ sin i
on a 58-d period orbit + a still heavier companion (m2 == 15.1 MJup/ sin i) further
out, orbits HD 168443. The inner planet was announced by Marcy et al. (1999b)
but with a much smaller mass (m2 == 4 MJup/ sin i). Their RVs combined with our
CORALIE observations allow us to simultaneously derive Keplerian solutions for
the 2 companions. Stellar and orbital parameters, inferred companion characteristics and RV curves are given in Table 1 and Fig. 2. The outer period is slightly
Downloaded from https:/www.cambridge.org/core. IP address: 88.99.165.207, on 18 Jun 2017 at 00:26:26, subject to the Cambridge Core terms of
use, available at https:/www.cambridge.org/core/terms. https://doi.org/10.1017/S0074180900217403
CORALIE-ELODIE new planets and planetary systems
17
larger than 1660 days. The period has not been completely covered yet but the
orbital solution is already very robust, the minimum and maximum phases of
the curve being covered. As for HD 83443 a preliminary stability study has been
performed for the system by W. Benz (Fig. 2, lower right). Strong dynamical
interactions between the components on various time scales (some related to the
orbital periods) are clearly visible on the eccentricity and separation temporal
variations. In particular, the large eccentricity value measured for the inner orbit
(e == 0.52) is very probably due to the gravitational perturbation of the heavier
outer companion. The system remains however stable.
The properties of HD 168443, if not due to an unlikely orbital plane inclination for both objects, set very interesting questions on the possible formation
of such systems. The massive inner planet and the low-mass brown-dwarf companion are located well within the ice limit of the young protoplanetary disk.
According to the migration scenario both objects had thus to move simultaneously towards the central region of the system. Simulations by Kley (2000)
explore this possibility and show that in such a case, the outer object accreates
disk material more efficiently. This rises the possibility of creating in the disk
"super planets" with masses larger than 15 MJup. On the other hand, if not in
the disk the more massive outer companion is close enough to have truncated
the disk within the ice limit thus preventing the giant planet to form.
4.
4.1.
Orbital element distributions: fossil traces of planet formation
The mass function of substellar companions
The observed gap in the mass distribution between giant planets and stellar
secondaries (Fig. 3) strongly suggests the existence of two distinct companion
populations for solar-type stars. The huge planetary peak is not the tail of the
binary distribution. Despite the huge observational bias against the detection
of small mass companions we observe an increasing number of low-mass planets.
The planetary mass function even increases towards the lower masses. In the
same time, the easier-detected brown dwarfs are rare. This is further emphasized
when orbital plane inclinations can be derived e.g. with Hipparcos astrometric
data (Halbwachs et al. 2000, Zucker & Mazeh 2000), transit observations or
synchronisation considerations.
The sharp drop of the mass function observed around 8 MJup strongly suggests a maximum mass for giant planets close to 10 M Jup. In particular, the
shape of the mass function does not suggest a relation between the
D-burning limit and the maximum mass for giant planets".
4.2.
Short-period and eccentricity distributions
Due to the still strong bias affecting the detection of long-period planets, a
significant comparison between planetary candidates and spectroscopic binaries
should be restricted to short-period systems. Figure 4 (left) focuses on P ::;
IThe discovery of "free-floating brown dwarfs" in a-Orionis with masses probably below 10 MJup
(Zapatero Osorio et al. 2000) further refutes the D-burning limit as a good indicator of the
brown dwarf - planet transition
Downloaded from https:/www.cambridge.org/core. IP address: 88.99.165.207, on 18 Jun 2017 at 00:26:26, subject to the Cambridge Core terms of
use, available at https:/www.cambridge.org/core/terms. https://doi.org/10.1017/S0074180900217403
18
Udry, Mayor and Queloz
Figure 3.
Mass function of companions to solar-type stars, including our
latest detections, in log (left) and linear (right) scales. The dotted vertical
lines indicate the H- and D-burning limits. Top: m2 sin i. Bottom: composite
histograms of m2 (open part) and m2 sin i (if sin i not known; colored part)
10 days, comparing the corresponding cumulative distributions. The planetary
period distribution is steeply rising for decreasing periods down to about 3 days
reaching a clear cut-off. Several reasons may be invoked to set this limit like
e.g. central cavity of the accretion disk, tidal interaction, Roche lobe overflow
or evaporation.
Concerning orbital eccentricities, the comparison between spectroscopic binaries and giant planets is remarkable (an (e, log P) comparison diagram is
given in Mayor & Udry 2000). No obvious difference appears except in particular domains of periods. For periods larger than about 50 days, we cannot
see any significant distinctions between both populations (Fig. 4, right) even if
the origin of the eccentricity is supposed to be different (e.g. Mayor & Udry
2000 for a further discussion). Nevertheless, a clear difference may be noticed
between the two populations for periods between 10 and 50 days (Fig. 4, right).
In this range of periods, outside the circularization domain, planetary systems
have smaller eccentricities than stellar binaries indicating different formations or
evolutions. Quasi-circular long-period binary orbits are very rare. Despite the
observational bias a similar situation already exists for extra-solar giant planets.
No solar-system analog has been discovered yet.
'l:0.8
I::
o
~0.6
>
~0.4
1
0.2
4
6
Period [days]
10
o
0.2 0.4 0.6 0.8
e
1 0
0.2 0.4 0.6 0.8
e
1
Figure 4.
Cumulative distributions for planetary (solid line) and stellar
companions (dashed line) to solar-type stars of: periods smaller than 10 days
(left) and eccentricities for two domains of periods (right)
Downloaded from https:/www.cambridge.org/core. IP address: 88.99.165.207, on 18 Jun 2017 at 00:26:26, subject to the Cambridge Core terms of
use, available at https:/www.cambridge.org/core/terms. https://doi.org/10.1017/S0074180900217403
CORALIE-ELODIE new planets and planetary systems
19
In conclusion, the differences observed between planetary systems and binaries in the metallicity content of the primaries (see e.g. Santos, Isrealian &
Mayor, this volume; Santos et al. 2000), in the secondary mass functions and
in the period and eccentricity distributions argue for the two populations being formed by distinct processes. More observations are however still needed to
bring clear constraints on the possible formation and evolution scenarios.
References
Baranne A., Queloz D., Mayor M. et al. 1996, A&AS, 119, 373
Charbonneau D., Brown T., Latham D., & Mayor M. 2000, ApJ, 529, L45
Delfosse X., Forveille T., Mayor M. et al. 1998, A&A, 338, L67
ESO 2000, http://www.eso.org/outreach/press-rel/pr-2000/pr-13-00.html
Halbwachs J.-L., Arenou F., Mayor M., Udry S., & Queloz D. 2000, A&A, 355,581
Henry G., Marcy G., Butler P., & Vogt S. 2000, ApJ, 529, L41
Kley W. 2000, MNRAS, 313, L47
Mazeh T., Naef D., Torres G. et al. 2000, ApJ, 532, L55
Marcy G., Butler P., Vogt S., Fischer D., & Lissauer J. 1999a, ApJ, 505, L147
Marcy G., Butler P., Vogt S., Fischer D., & Liu M. 1999b, ApJ, 520,239
Mayor M., & Queloz D. 1995, Nature, 378, 355
Mayor M., & Udry S. 2000, in "Disks, planetesimals and planets", Tenerife, ASP Conf.
Ser., in press
Naef D., Mayor M., Pepe F. et al. 2000b, A&A, in press
Queloz D., Eggenberger A., Mayor M. et al. 2000, A&A, 359, L13
Santos N.C., Israelian G., & Mayor M. 2000, A&A, in press
Udry S., Mayor M., & Naef D., et al. 2000, A&A, 356, 590
Zapatero Osorio M.-R., Bejar V., Martin E. et al. 2000, Science, 290, 103
.Zucker S., & Mazeh T. 2000, ApJ, 531, L67
Downloaded from https:/www.cambridge.org/core. IP address: 88.99.165.207, on 18 Jun 2017 at 00:26:26, subject to the Cambridge Core terms of
use, available at https:/www.cambridge.org/core/terms. https://doi.org/10.1017/S0074180900217403